Clinical Inquiries

Evidence summary

Multiple analyses suggest HRT worsens rather than improves cognition

A 2017 Cochrane review of 22 randomized, double-blind studies compared use of HRT (estrogen only or combination estrogen + progesterone therapies) with placebo in postmenopausal women (N = 43,637). Age ranges varied, but the average age in most studies was > 60 years. Treatment duration was at least 1 year. Various outcomes were assessed across these 22 studies, including cardiovascular disease, bone health, and cognition.¹

Cognitive outcomes were assessed with the Mini-Mental Status Exam in 5 of the trials (N = 12,789). Data were not combined due to heterogeneity. The authors found no significant difference in cognitive scores between the treatment and control groups in any of these 5 studies.¹

In the largest included study, the Women’s Health Initiative (WHI) Memory Study (N = 10,739), participants were older than 65 years. Among those receiving estrogen-only HRT, there were no statistically significant differences compared to those receiving placebo. However, healthy postmenopausal women taking combination HRT had an increased risk for “probable dementia” compared to those taking placebo (relative risk [RR] = 1.97; 95% CI, 1.16-3.33). When researchers looked exclusively at women taking HRT, the risk for dementia increased from 9 in 1000 to 18 in 1000 (95% CI, 11-30) after 4 years of HRT use. This results in a number needed to harm of 4 to 50 patients.¹

Two notable limitations of this evidence are that the average age of this population was > 60 years and 80% of the participants were White.¹

A 2021 meta-analysis of 23 RCTs (N = 13,683) studied the effect of HRT on global cognitive function as well as specific cognitive domains including memory, executive function, attention, and language. Mean patient age in the studies varied from 48 to 81 years. Nine of these studies were also included in the previously discussed Cochrane review.²

There was a statistically significant but small decrease in overall global cognition (10 trials; N = 12,115; standardized mean difference [SMD] = –0.04; 95% CI, –0.08 to –0.01) in those receiving HRT compared to placebo. This effect was slightly more pronounced among those who initiated HRT at age > 60 years (8 trials; N = 11,914; SMD = –0.05; 95% CI, –0.08 to –0.01) and among patients with HRT duration > 6 months (7 trials; N = 11,828; SMD = –0.05; 95% CI, –0.08 to –0.01). There were no significant differences in specific cognitive domains.²

In a 2017 follow-up to the WHI trial, researchers analyzed data on long-term cognitive effects in women previously treated with HRT. There were 2 cohorts: participants who initiated HRT at a younger age (50-54; N = 1376) and those who initiated HRT later in life (age 65-79; N = 2880). Cognitive outcomes were assessed using the Telephone Interview for Cognitive Status-modified, with interviews conducted annually beginning 6 to 7 years after HRT was stopped.³

The investigators found no significant change in composite cognitive function in the younger HRT-treated group compared to placebo (estrogen alone: mean deviation [MD] = 0.014; 95% CI, –0.097 to 0.126; estrogen + progesterone: MD = –0.047; 95% CI, –0.134 to 0.04), or in the group who initiated HRT at an older age (estrogen alone: MD = –0.099; 95% CI, –0.202 to 0.004; estrogen + progesterone: MD = –0.022; 95% CI, –0.099 to 0.055). The authors state that although the data did not reach significance, this study also found a trend toward decreases in global cognitive function in the older age group.³

Editor’s takeaway

Abundant, consistent evidence with long-term follow-up shows postmenopausal HRT does not reduce cognitive decline. In fact, it appears to increase cognitive decline slightly. Renewed interest in postmenopausal HRT to alleviate menopausal symptoms should balance the risks and benefits to the individual patient.

Evidence summary

Multiple analyses suggest HRT worsens rather than improves cognition

A 2017 Cochrane review of 22 randomized, double-blind studies compared use of HRT (estrogen only or combination estrogen + progesterone therapies) with placebo in postmenopausal women (N = 43,637). Age ranges varied, but the average age in most studies was > 60 years. Treatment duration was at least 1 year. Various outcomes were assessed across these 22 studies, including cardiovascular disease, bone health, and cognition.¹

Cognitive outcomes were assessed with the Mini-Mental Status Exam in 5 of the trials (N = 12,789). Data were not combined due to heterogeneity. The authors found no significant difference in cognitive scores between the treatment and control groups in any of these 5 studies.¹

In the largest included study, the Women’s Health Initiative (WHI) Memory Study (N = 10,739), participants were older than 65 years. Among those receiving estrogen-only HRT, there were no statistically significant differences compared to those receiving placebo. However, healthy postmenopausal women taking combination HRT had an increased risk for “probable dementia” compared to those taking placebo (relative risk [RR] = 1.97; 95% CI, 1.16-3.33). When researchers looked exclusively at women taking HRT, the risk for dementia increased from 9 in 1000 to 18 in 1000 (95% CI, 11-30) after 4 years of HRT use. This results in a number needed to harm of 4 to 50 patients.¹

Two notable limitations of this evidence are that the average age of this population was > 60 years and 80% of the participants were White.¹

A 2021 meta-analysis of 23 RCTs (N = 13,683) studied the effect of HRT on global cognitive function as well as specific cognitive domains including memory, executive function, attention, and language. Mean patient age in the studies varied from 48 to 81 years. Nine of these studies were also included in the previously discussed Cochrane review.²

There was a statistically significant but small decrease in overall global cognition (10 trials; N = 12,115; standardized mean difference [SMD] = –0.04; 95% CI, –0.08 to –0.01) in those receiving HRT compared to placebo. This effect was slightly more pronounced among those who initiated HRT at age > 60 years (8 trials; N = 11,914; SMD = –0.05; 95% CI, –0.08 to –0.01) and among patients with HRT duration > 6 months (7 trials; N = 11,828; SMD = –0.05; 95% CI, –0.08 to –0.01). There were no significant differences in specific cognitive domains.²

In a 2017 follow-up to the WHI trial, researchers analyzed data on long-term cognitive effects in women previously treated with HRT. There were 2 cohorts: participants who initiated HRT at a younger age (50-54; N = 1376) and those who initiated HRT later in life (age 65-79; N = 2880). Cognitive outcomes were assessed using the Telephone Interview for Cognitive Status-modified, with interviews conducted annually beginning 6 to 7 years after HRT was stopped.³

The investigators found no significant change in composite cognitive function in the younger HRT-treated group compared to placebo (estrogen alone: mean deviation [MD] = 0.014; 95% CI, –0.097 to 0.126; estrogen + progesterone: MD = –0.047; 95% CI, –0.134 to 0.04), or in the group who initiated HRT at an older age (estrogen alone: MD = –0.099; 95% CI, –0.202 to 0.004; estrogen + progesterone: MD = –0.022; 95% CI, –0.099 to 0.055). The authors state that although the data did not reach significance, this study also found a trend toward decreases in global cognitive function in the older age group.³

Editor’s takeaway

Abundant, consistent evidence with long-term follow-up shows postmenopausal HRT does not reduce cognitive decline. In fact, it appears to increase cognitive decline slightly. Renewed interest in postmenopausal HRT to alleviate menopausal symptoms should balance the risks and benefits to the individual patient.

Evidence summary

Multiple analyses suggest HRT worsens rather than improves cognition

A 2017 Cochrane review of 22 randomized, double-blind studies compared use of HRT (estrogen only or combination estrogen + progesterone therapies) with placebo in postmenopausal women (N = 43,637). Age ranges varied, but the average age in most studies was > 60 years. Treatment duration was at least 1 year. Various outcomes were assessed across these 22 studies, including cardiovascular disease, bone health, and cognition.¹

Cognitive outcomes were assessed with the Mini-Mental Status Exam in 5 of the trials (N = 12,789). Data were not combined due to heterogeneity. The authors found no significant difference in cognitive scores between the treatment and control groups in any of these 5 studies.¹

In the largest included study, the Women’s Health Initiative (WHI) Memory Study (N = 10,739), participants were older than 65 years. Among those receiving estrogen-only HRT, there were no statistically significant differences compared to those receiving placebo. However, healthy postmenopausal women taking combination HRT had an increased risk for “probable dementia” compared to those taking placebo (relative risk [RR] = 1.97; 95% CI, 1.16-3.33). When researchers looked exclusively at women taking HRT, the risk for dementia increased from 9 in 1000 to 18 in 1000 (95% CI, 11-30) after 4 years of HRT use. This results in a number needed to harm of 4 to 50 patients.¹

Two notable limitations of this evidence are that the average age of this population was > 60 years and 80% of the participants were White.¹

A 2021 meta-analysis of 23 RCTs (N = 13,683) studied the effect of HRT on global cognitive function as well as specific cognitive domains including memory, executive function, attention, and language. Mean patient age in the studies varied from 48 to 81 years. Nine of these studies were also included in the previously discussed Cochrane review.²

There was a statistically significant but small decrease in overall global cognition (10 trials; N = 12,115; standardized mean difference [SMD] = –0.04; 95% CI, –0.08 to –0.01) in those receiving HRT compared to placebo. This effect was slightly more pronounced among those who initiated HRT at age > 60 years (8 trials; N = 11,914; SMD = –0.05; 95% CI, –0.08 to –0.01) and among patients with HRT duration > 6 months (7 trials; N = 11,828; SMD = –0.05; 95% CI, –0.08 to –0.01). There were no significant differences in specific cognitive domains.²

In a 2017 follow-up to the WHI trial, researchers analyzed data on long-term cognitive effects in women previously treated with HRT. There were 2 cohorts: participants who initiated HRT at a younger age (50-54; N = 1376) and those who initiated HRT later in life (age 65-79; N = 2880). Cognitive outcomes were assessed using the Telephone Interview for Cognitive Status-modified, with interviews conducted annually beginning 6 to 7 years after HRT was stopped.³

The investigators found no significant change in composite cognitive function in the younger HRT-treated group compared to placebo (estrogen alone: mean deviation [MD] = 0.014; 95% CI, –0.097 to 0.126; estrogen + progesterone: MD = –0.047; 95% CI, –0.134 to 0.04), or in the group who initiated HRT at an older age (estrogen alone: MD = –0.099; 95% CI, –0.202 to 0.004; estrogen + progesterone: MD = –0.022; 95% CI, –0.099 to 0.055). The authors state that although the data did not reach significance, this study also found a trend toward decreases in global cognitive function in the older age group.³

Editor’s takeaway

Abundant, consistent evidence with long-term follow-up shows postmenopausal HRT does not reduce cognitive decline. In fact, it appears to increase cognitive decline slightly. Renewed interest in postmenopausal HRT to alleviate menopausal symptoms should balance the risks and benefits to the individual patient.

Evidence summary

Mixed results, Yes, but no evidence of harm in continuing RAAS therapy

A 2014 cohort study assessed the effect of treatment with angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEIs/ARBs) on all-cause mortality in US veterans (N = 141,413) with non-dialysis chronic kidney disease (CKD)—defined as either a stable estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m² or a stable eGFR ≥ 60 mL/min/1.73 m² and an elevated urine microalbumin measurement.¹ In an intention-to-treat analysis, ACEI/ARB treatment was associated with a significantly decreased risk for all-cause mortality (hazard ratio [HR] = 0.81; 95% CI, 0.78-0.84).

A 2018 meta-analysis analyzed data from 9 RCTs comparing RAAS blockade therapy to placebo or alternative antihypertensive agents in patients with non-dialysis CKD stages 3 to 5.² Although the meta-analysis authors focused on patients with comorbid diabetes and non-dialysis CKD (N = 9797), some included studies had a mixed population (ie, only a subset of patients had diabetes). This, among other variances in characteristics, participants, interventions, and endpoints, resulted in different numbers of participants included in the data extraction and analysis of outcomes. Overall, there was no difference between the RAAS group and the control group in terms of all-cause mortality (N = 5309; risk ratio [RR] = 0.97; 95% CI, 0.85-1.10), cardiovascular mortality (N = 3748; RR = 1.03; 95% CI, 0.75-1.41), or adverse events (N = 1822; RR = 1.05; 95% CI, 0.89-1.25). Compared to the control group, the RAAS group was less likely to experience a nonfatal cardiovascular event (N = 6138; RR = 0.90; 95% CI, 0.81-1.00). For the composite endpoint of need for renal replacement therapy/doubling of serum creatinine, RAAS therapy was associated with reduced risk in both the overall population (N = 5202; RR = 0.81; 95% CI, 0.70-0.92) and in patients with comorbid diabetes (N = 3314; RR = 0.78; 95% CI, 0.67-0.90).

A 2022 open-label trial (STOP ACEi) randomly assigned 411 patients with stage 4 or 5 CKD to either continue (N = 205) or discontinue (N = 206) RAAS inhibitor therapy.³ The primary outcome measure was eGFR at 3 years. The difference in the rate of decline in eGFR between groups was –0.7% (95% CI, –2.5 to 1.0; P = .42), favoring the group that continued therapy.

Recommendations from others

After reviewing data from multiple clinical trials, the authors of the 2018 report from the National Kidney Foundation–Kidney Disease Outcomes Quality Initiative (NKF–KDOQI) concluded that the decision to continue or stop RAAS therapy in patients with advanced CKD should be individualized.⁴ Criteria that should be considered in the decision-making process include the presence or absence of large acute declines in eGFR (> 20% in the absence of a significant decrease in proteinuria), hypotension, or acute kidney injury with significant risk for worsening.

In 2021, the Renal Association and the Association of British Clinical Diabetologists published updated clinical practice guidelines for the management of hypertension and RAAS blockade in adults with diabetic kidney disease.⁵ Collective data indicated that, although outcomes varied based on type of diabetes (1 vs 2) and degree of proteinuria, blockade therapy overall led to improved outcomes; this was hypothesized to be due to the effects of reduced blood pressure. However, discontinuation of RAAS blockade therapy may be warranted when the patient (1) has a potassium level > 5 mmol/L pretreatment or ≥ 6 mmol/L with treatment, (2) demonstrates a decrease in eGFR > 25% or an increase in serum creatinine > 30% upon initiation of blockade, without another cause of renal deterioration, (3) is pregnant, or (4) has an acute illness with fluid depletion (in which case, RAAS therapy can be restarted 24 to 48 hours after recovery).

Editor’s takeaway

Evidence supports continuation of RAAS blockade, particularly in patients with significant comorbidities (diabetes and cardiovascular disease). Study data indicate continuation is either beneficial or neutral to further morbidity. The only caveat is that these patients should have their renal function and potassium level continuously monitored. The evidence should provide reassurance to patients and physicians that continuation is the correct course of action.

Evidence summary

Mixed results, Yes, but no evidence of harm in continuing RAAS therapy

A 2014 cohort study assessed the effect of treatment with angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEIs/ARBs) on all-cause mortality in US veterans (N = 141,413) with non-dialysis chronic kidney disease (CKD)—defined as either a stable estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m² or a stable eGFR ≥ 60 mL/min/1.73 m² and an elevated urine microalbumin measurement.¹ In an intention-to-treat analysis, ACEI/ARB treatment was associated with a significantly decreased risk for all-cause mortality (hazard ratio [HR] = 0.81; 95% CI, 0.78-0.84).

A 2018 meta-analysis analyzed data from 9 RCTs comparing RAAS blockade therapy to placebo or alternative antihypertensive agents in patients with non-dialysis CKD stages 3 to 5.² Although the meta-analysis authors focused on patients with comorbid diabetes and non-dialysis CKD (N = 9797), some included studies had a mixed population (ie, only a subset of patients had diabetes). This, among other variances in characteristics, participants, interventions, and endpoints, resulted in different numbers of participants included in the data extraction and analysis of outcomes. Overall, there was no difference between the RAAS group and the control group in terms of all-cause mortality (N = 5309; risk ratio [RR] = 0.97; 95% CI, 0.85-1.10), cardiovascular mortality (N = 3748; RR = 1.03; 95% CI, 0.75-1.41), or adverse events (N = 1822; RR = 1.05; 95% CI, 0.89-1.25). Compared to the control group, the RAAS group was less likely to experience a nonfatal cardiovascular event (N = 6138; RR = 0.90; 95% CI, 0.81-1.00). For the composite endpoint of need for renal replacement therapy/doubling of serum creatinine, RAAS therapy was associated with reduced risk in both the overall population (N = 5202; RR = 0.81; 95% CI, 0.70-0.92) and in patients with comorbid diabetes (N = 3314; RR = 0.78; 95% CI, 0.67-0.90).

A 2022 open-label trial (STOP ACEi) randomly assigned 411 patients with stage 4 or 5 CKD to either continue (N = 205) or discontinue (N = 206) RAAS inhibitor therapy.³ The primary outcome measure was eGFR at 3 years. The difference in the rate of decline in eGFR between groups was –0.7% (95% CI, –2.5 to 1.0; P = .42), favoring the group that continued therapy.

Recommendations from others

After reviewing data from multiple clinical trials, the authors of the 2018 report from the National Kidney Foundation–Kidney Disease Outcomes Quality Initiative (NKF–KDOQI) concluded that the decision to continue or stop RAAS therapy in patients with advanced CKD should be individualized.⁴ Criteria that should be considered in the decision-making process include the presence or absence of large acute declines in eGFR (> 20% in the absence of a significant decrease in proteinuria), hypotension, or acute kidney injury with significant risk for worsening.

In 2021, the Renal Association and the Association of British Clinical Diabetologists published updated clinical practice guidelines for the management of hypertension and RAAS blockade in adults with diabetic kidney disease.⁵ Collective data indicated that, although outcomes varied based on type of diabetes (1 vs 2) and degree of proteinuria, blockade therapy overall led to improved outcomes; this was hypothesized to be due to the effects of reduced blood pressure. However, discontinuation of RAAS blockade therapy may be warranted when the patient (1) has a potassium level > 5 mmol/L pretreatment or ≥ 6 mmol/L with treatment, (2) demonstrates a decrease in eGFR > 25% or an increase in serum creatinine > 30% upon initiation of blockade, without another cause of renal deterioration, (3) is pregnant, or (4) has an acute illness with fluid depletion (in which case, RAAS therapy can be restarted 24 to 48 hours after recovery).

Editor’s takeaway

Evidence supports continuation of RAAS blockade, particularly in patients with significant comorbidities (diabetes and cardiovascular disease). Study data indicate continuation is either beneficial or neutral to further morbidity. The only caveat is that these patients should have their renal function and potassium level continuously monitored. The evidence should provide reassurance to patients and physicians that continuation is the correct course of action.

Evidence summary

Mixed results, Yes, but no evidence of harm in continuing RAAS therapy

A 2014 cohort study assessed the effect of treatment with angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEIs/ARBs) on all-cause mortality in US veterans (N = 141,413) with non-dialysis chronic kidney disease (CKD)—defined as either a stable estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m² or a stable eGFR ≥ 60 mL/min/1.73 m² and an elevated urine microalbumin measurement.¹ In an intention-to-treat analysis, ACEI/ARB treatment was associated with a significantly decreased risk for all-cause mortality (hazard ratio [HR] = 0.81; 95% CI, 0.78-0.84).

A 2018 meta-analysis analyzed data from 9 RCTs comparing RAAS blockade therapy to placebo or alternative antihypertensive agents in patients with non-dialysis CKD stages 3 to 5.² Although the meta-analysis authors focused on patients with comorbid diabetes and non-dialysis CKD (N = 9797), some included studies had a mixed population (ie, only a subset of patients had diabetes). This, among other variances in characteristics, participants, interventions, and endpoints, resulted in different numbers of participants included in the data extraction and analysis of outcomes. Overall, there was no difference between the RAAS group and the control group in terms of all-cause mortality (N = 5309; risk ratio [RR] = 0.97; 95% CI, 0.85-1.10), cardiovascular mortality (N = 3748; RR = 1.03; 95% CI, 0.75-1.41), or adverse events (N = 1822; RR = 1.05; 95% CI, 0.89-1.25). Compared to the control group, the RAAS group was less likely to experience a nonfatal cardiovascular event (N = 6138; RR = 0.90; 95% CI, 0.81-1.00). For the composite endpoint of need for renal replacement therapy/doubling of serum creatinine, RAAS therapy was associated with reduced risk in both the overall population (N = 5202; RR = 0.81; 95% CI, 0.70-0.92) and in patients with comorbid diabetes (N = 3314; RR = 0.78; 95% CI, 0.67-0.90).

A 2022 open-label trial (STOP ACEi) randomly assigned 411 patients with stage 4 or 5 CKD to either continue (N = 205) or discontinue (N = 206) RAAS inhibitor therapy.³ The primary outcome measure was eGFR at 3 years. The difference in the rate of decline in eGFR between groups was –0.7% (95% CI, –2.5 to 1.0; P = .42), favoring the group that continued therapy.

Recommendations from others

After reviewing data from multiple clinical trials, the authors of the 2018 report from the National Kidney Foundation–Kidney Disease Outcomes Quality Initiative (NKF–KDOQI) concluded that the decision to continue or stop RAAS therapy in patients with advanced CKD should be individualized.⁴ Criteria that should be considered in the decision-making process include the presence or absence of large acute declines in eGFR (> 20% in the absence of a significant decrease in proteinuria), hypotension, or acute kidney injury with significant risk for worsening.

In 2021, the Renal Association and the Association of British Clinical Diabetologists published updated clinical practice guidelines for the management of hypertension and RAAS blockade in adults with diabetic kidney disease.⁵ Collective data indicated that, although outcomes varied based on type of diabetes (1 vs 2) and degree of proteinuria, blockade therapy overall led to improved outcomes; this was hypothesized to be due to the effects of reduced blood pressure. However, discontinuation of RAAS blockade therapy may be warranted when the patient (1) has a potassium level > 5 mmol/L pretreatment or ≥ 6 mmol/L with treatment, (2) demonstrates a decrease in eGFR > 25% or an increase in serum creatinine > 30% upon initiation of blockade, without another cause of renal deterioration, (3) is pregnant, or (4) has an acute illness with fluid depletion (in which case, RAAS therapy can be restarted 24 to 48 hours after recovery).

Editor’s takeaway

Evidence supports continuation of RAAS blockade, particularly in patients with significant comorbidities (diabetes and cardiovascular disease). Study data indicate continuation is either beneficial or neutral to further morbidity. The only caveat is that these patients should have their renal function and potassium level continuously monitored. The evidence should provide reassurance to patients and physicians that continuation is the correct course of action.

Evidence summary

Association between higher maternal BMI and increased risk for stillbirth

The purpose of antenatal testing is to decrease the risk for stillbirth between visits. Because of the resources involved and the risk for false-positives when testing low-risk patients, antenatal testing is reserved for pregnant people with higher risk for stillbirth.

In a retrospective cohort study of more than 2.8 million singleton births including 9030 stillbirths, pregnant people with an ­elevated BMI had an increased risk for stillbirth compared to those with a normal BMI. The adjusted hazard ratio was 1.71 (95% CI, 1.62-1.83) for those with a BMI of 30.0 to 34.9; 2.04 (95% CI, 1.8-2.21) for those with a BMI of 35.0 to 39.9; and 2.50 (95% CI, 2.28-2.74) for those with a BMI ≥ 40.¹

A meta-analysis of 38 studies, which included data on 16,274 stillbirths, found that a 5-unit increase in BMI was associated with an increased risk for stillbirth (relative risk, 1.24; 95% CI, 1.18-1.30).²

Another meta-analysis included 6 cohort studies involving more than 1 million pregnancies and 3 case-control studies involving 2530 stillbirths and 2837 controls from 1980-2005. There was an association between increasing BMI and stillbirth: the odds ratio (OR) was 1.47 (95% CI, 1.08-1.94) for those with a BMI of 25.0 to 29.9 and 2.07 (95% CI, 1.59-2.74) for those with a BMI ≥ 30.0, compared to those with a normal BMI.³

However, a retrospective cohort study of 182,362 singleton births including 442 stillbirths found no association between stillbirth and increasing BMI. The OR was 1.10 (95% CI, 0.90-1.36) for those with a BMI of 25.0 to 29.9 and 1.09 (95% CI, 0.87-1.37) for those with a BMI ≥ 30.0, compared to those with a normal BMI.⁴ However, this cohort study may have been underpowered to detect an association between stillbirth and BMI.

Recommendations from others

In 2021, ACOG suggested that weekly antenatal testing may be considered from 34w0d for pregnant people with a BMI ≥ 40.0 and from 37w0d for pregnant people with a BMI between 35.0 and 39.9.⁵ The 2021 ACOG Practice Bulletin on Obesity in Pregnancy rates this recommendation as Level C—based primarily on consensus and expert opinion.⁶

A 2018 Royal College of Obstetricians and Gynecologists Green-top Guideline recognizes “definitive recommendations for fetal surveillance are hampered by the lack of randomized controlled trials demonstrating that antepartum fetal surveillance decreases perinatal morbidity or mortality in late-term and post-term gestations…. There are no definitive studies determining the optimal type or frequency of such testing and no evidence specific for women with obesity.”⁷

A 2019 Society of Obstetricians and Gynecologists of Canada practice guideline states “stillbirth is more common with maternal obesity” and recommends “increased fetal surveillance … in the third trimester if reduced fetal movements are reported.” The guideline notes “the role for non-stress tests … in surveillance of well-being in this population is uncertain.” Also, for pregnant people with a BMI > 30, “assessment of fetal well-being is … recommended weekly from 37 weeks until delivery.” Finally, increased fetal surveillance is recommended in the setting of increased BMI and an abnormal pulsatility index of the umbilical artery and/or maternal uterine artery.⁸

Editor’s takeaway

Evidence demonstrates that increased maternal BMI is associated with increased stillbirths. However, evidence has not shown that third-trimester antenatal testing decreases this morbidity and mortality. Expert opinion varies, with ACOG recommending weekly antenatal testing from 34 and 37 weeks for pregnant people with a BMI ≥ 40 and of 35 to 39.9, respectively.

Evidence summary

Association between higher maternal BMI and increased risk for stillbirth

The purpose of antenatal testing is to decrease the risk for stillbirth between visits. Because of the resources involved and the risk for false-positives when testing low-risk patients, antenatal testing is reserved for pregnant people with higher risk for stillbirth.

In a retrospective cohort study of more than 2.8 million singleton births including 9030 stillbirths, pregnant people with an ­elevated BMI had an increased risk for stillbirth compared to those with a normal BMI. The adjusted hazard ratio was 1.71 (95% CI, 1.62-1.83) for those with a BMI of 30.0 to 34.9; 2.04 (95% CI, 1.8-2.21) for those with a BMI of 35.0 to 39.9; and 2.50 (95% CI, 2.28-2.74) for those with a BMI ≥ 40.¹

A meta-analysis of 38 studies, which included data on 16,274 stillbirths, found that a 5-unit increase in BMI was associated with an increased risk for stillbirth (relative risk, 1.24; 95% CI, 1.18-1.30).²

Another meta-analysis included 6 cohort studies involving more than 1 million pregnancies and 3 case-control studies involving 2530 stillbirths and 2837 controls from 1980-2005. There was an association between increasing BMI and stillbirth: the odds ratio (OR) was 1.47 (95% CI, 1.08-1.94) for those with a BMI of 25.0 to 29.9 and 2.07 (95% CI, 1.59-2.74) for those with a BMI ≥ 30.0, compared to those with a normal BMI.³

However, a retrospective cohort study of 182,362 singleton births including 442 stillbirths found no association between stillbirth and increasing BMI. The OR was 1.10 (95% CI, 0.90-1.36) for those with a BMI of 25.0 to 29.9 and 1.09 (95% CI, 0.87-1.37) for those with a BMI ≥ 30.0, compared to those with a normal BMI.⁴ However, this cohort study may have been underpowered to detect an association between stillbirth and BMI.

Recommendations from others

In 2021, ACOG suggested that weekly antenatal testing may be considered from 34w0d for pregnant people with a BMI ≥ 40.0 and from 37w0d for pregnant people with a BMI between 35.0 and 39.9.⁵ The 2021 ACOG Practice Bulletin on Obesity in Pregnancy rates this recommendation as Level C—based primarily on consensus and expert opinion.⁶

A 2018 Royal College of Obstetricians and Gynecologists Green-top Guideline recognizes “definitive recommendations for fetal surveillance are hampered by the lack of randomized controlled trials demonstrating that antepartum fetal surveillance decreases perinatal morbidity or mortality in late-term and post-term gestations…. There are no definitive studies determining the optimal type or frequency of such testing and no evidence specific for women with obesity.”⁷

A 2019 Society of Obstetricians and Gynecologists of Canada practice guideline states “stillbirth is more common with maternal obesity” and recommends “increased fetal surveillance … in the third trimester if reduced fetal movements are reported.” The guideline notes “the role for non-stress tests … in surveillance of well-being in this population is uncertain.” Also, for pregnant people with a BMI > 30, “assessment of fetal well-being is … recommended weekly from 37 weeks until delivery.” Finally, increased fetal surveillance is recommended in the setting of increased BMI and an abnormal pulsatility index of the umbilical artery and/or maternal uterine artery.⁸

Editor’s takeaway

Evidence demonstrates that increased maternal BMI is associated with increased stillbirths. However, evidence has not shown that third-trimester antenatal testing decreases this morbidity and mortality. Expert opinion varies, with ACOG recommending weekly antenatal testing from 34 and 37 weeks for pregnant people with a BMI ≥ 40 and of 35 to 39.9, respectively.

Evidence summary

Association between higher maternal BMI and increased risk for stillbirth

The purpose of antenatal testing is to decrease the risk for stillbirth between visits. Because of the resources involved and the risk for false-positives when testing low-risk patients, antenatal testing is reserved for pregnant people with higher risk for stillbirth.

In a retrospective cohort study of more than 2.8 million singleton births including 9030 stillbirths, pregnant people with an ­elevated BMI had an increased risk for stillbirth compared to those with a normal BMI. The adjusted hazard ratio was 1.71 (95% CI, 1.62-1.83) for those with a BMI of 30.0 to 34.9; 2.04 (95% CI, 1.8-2.21) for those with a BMI of 35.0 to 39.9; and 2.50 (95% CI, 2.28-2.74) for those with a BMI ≥ 40.¹

A meta-analysis of 38 studies, which included data on 16,274 stillbirths, found that a 5-unit increase in BMI was associated with an increased risk for stillbirth (relative risk, 1.24; 95% CI, 1.18-1.30).²

Another meta-analysis included 6 cohort studies involving more than 1 million pregnancies and 3 case-control studies involving 2530 stillbirths and 2837 controls from 1980-2005. There was an association between increasing BMI and stillbirth: the odds ratio (OR) was 1.47 (95% CI, 1.08-1.94) for those with a BMI of 25.0 to 29.9 and 2.07 (95% CI, 1.59-2.74) for those with a BMI ≥ 30.0, compared to those with a normal BMI.³

However, a retrospective cohort study of 182,362 singleton births including 442 stillbirths found no association between stillbirth and increasing BMI. The OR was 1.10 (95% CI, 0.90-1.36) for those with a BMI of 25.0 to 29.9 and 1.09 (95% CI, 0.87-1.37) for those with a BMI ≥ 30.0, compared to those with a normal BMI.⁴ However, this cohort study may have been underpowered to detect an association between stillbirth and BMI.

Recommendations from others

In 2021, ACOG suggested that weekly antenatal testing may be considered from 34w0d for pregnant people with a BMI ≥ 40.0 and from 37w0d for pregnant people with a BMI between 35.0 and 39.9.⁵ The 2021 ACOG Practice Bulletin on Obesity in Pregnancy rates this recommendation as Level C—based primarily on consensus and expert opinion.⁶

A 2018 Royal College of Obstetricians and Gynecologists Green-top Guideline recognizes “definitive recommendations for fetal surveillance are hampered by the lack of randomized controlled trials demonstrating that antepartum fetal surveillance decreases perinatal morbidity or mortality in late-term and post-term gestations…. There are no definitive studies determining the optimal type or frequency of such testing and no evidence specific for women with obesity.”⁷

A 2019 Society of Obstetricians and Gynecologists of Canada practice guideline states “stillbirth is more common with maternal obesity” and recommends “increased fetal surveillance … in the third trimester if reduced fetal movements are reported.” The guideline notes “the role for non-stress tests … in surveillance of well-being in this population is uncertain.” Also, for pregnant people with a BMI > 30, “assessment of fetal well-being is … recommended weekly from 37 weeks until delivery.” Finally, increased fetal surveillance is recommended in the setting of increased BMI and an abnormal pulsatility index of the umbilical artery and/or maternal uterine artery.⁸

Editor’s takeaway

Evidence demonstrates that increased maternal BMI is associated with increased stillbirths. However, evidence has not shown that third-trimester antenatal testing decreases this morbidity and mortality. Expert opinion varies, with ACOG recommending weekly antenatal testing from 34 and 37 weeks for pregnant people with a BMI ≥ 40 and of 35 to 39.9, respectively.

Evidence summary

Walking’s impact on cholesterol levels is modest, inconsistent

A 2022 systematic review and meta-analysis of 21 studies (n = 1129) evaluated the effects of walking on lipids and lipoproteins in women older than 18 years who were overweight or obese and were not taking any lipid-­lowering medications. Median TC was 206 mg/dL and median LDL was 126 mg/dL.¹

The primary outcome found that walking decreased TC and LDL levels independent of diet and weight loss. Twenty studies reported on TC and showed that walking significantly decreased TC levels compared to the control groups (raw mean difference [RMD] = 6.7 mg/dL; 95% CI, 0.4-12.9; P = .04). Fifteen studies examined LDL and showed a significant decrease in LDL levels with walking compared to control groups (RMD = 7.4 mg/dL; 95% CI, 0.3-14.5; P = .04). However, the small magnitude of the changes may have little clinical impact.¹

There were no significant changes in the walking groups compared to the control groups for triglycerides (17 studies; RMD = 2.2 mg/dL; 95% CI, –8.4 to 12.8; P = .68) or high-density lipoprotein (HDL) (18 studies; RMD = 1.5 mg/dL; 95% CI, –0.4 to 3.3; P = .12). Included studies were required to be controlled but were otherwise not described. The overall risk for bias was determined to be low.¹

A 2020 RCT (n = 22) assessed the effects of a walking intervention on cholesterol and cardiovascular disease (CVD) risk in individuals ages 40 to 65 years with moderate CVD risk but without diabetes or CVD.² Moderate CVD risk was defined as a 2% to 5% ­10-year risk for a CVD event using the European HeartScore, which incorporates age, sex, blood pressure, lipid levels, and smoking status³; however, study participants were not required to have hyperlipidemia. Participants were enrolled in a 12-week, nurse-led intervention of moderate-paced walking for 30 to 45 minutes 5 times weekly.

Individuals in the intervention group had significant decreases in average TC levels from baseline to follow-up (244.6 mg/dL vs 213.7 mg/dL; P = .001). As a result, participants’ average 10-year CVD risk was significantly reduced from moderate risk to low risk (2.6% vs 1.8%; P = 038) and was significantly lower in the intervention group than in the control group at follow-up (1.8% vs 3.1%; P = .019). No blinding was used, and the use of lipid-lowering medications was not reported, which could have impacted the results.²

A 2008 RCT (n = 67) examined the effect of a home-based walking program (12 weeks of brisk walking, at least 30 min/d and at least 5 d/wk, with at least 300 kcal burned per walk) vs a sedentary control group in men ages 45 to 65 years with hyperlipidemia (TC > 240 mg/dL and/or TC/­HDL-C ratio ≥ 6) who were not receiving lipid-lowering medication. There were no significant changes from baseline to follow-up in the walking group compared to the control group in TC (adjusted mean difference [AMD] = –9.3 mg/dL; 95% CI, –22.8 to 4.64; P = .19), HDL-C (AMD = 2.7 mg/dL; 95% CI, –0.4 to 5.4; P = .07) or triglycerides (AMD = –26.6 mg/dL; 95% CI, –56.7 to 2.7; P = .07).⁴

The lipid reductions achieved from walking—if any—are minimal.

A 2002 RCT (n = 111) of sedentary men and women (BMI, 25-35; ages, 40-65 years) with dyslipidemia (LDL of 130-190 mg/dL, or HDL < 40 mg/dL for men or < 45 mg/dL for women) examined the impact of various physical activity levels for 8 months when compared to a control group observed for 6 months. The group assigned to low-amount, moderate-intensity physical activity walked an equivalent of 12 miles per week.⁵

Continue to: In this group...

In this group, there was a significant decrease in average triglyceride concentrations from baseline to follow-up (mean ± standard error = 196.8 ± 30.5 mg/dL vs 145.2 ± 16.0 mg/dL; P < .001). Significance of the change compared with changes in the control group was not reported, although triglycerides in the control group increased from baseline to follow-up (132.1 ± 11.0 vs 155.8 ± 14.9 mg/dL). There were no significant changes from baseline to follow-up in TC (194 ± 4.8 vs 197.9 ± 5.4 mg/dL), LDL (122.7 ± 4.0 vs 127.8 ± 4.1 mg/dL), or HDL (42.0 ± 1.9 vs 43.1 ± 2.5 mg/dL); P values of pre-post changes and comparison to control group were not reported.⁵

Recommendations from others

The Physical Activity Guidelines for Americans, published by the Department of Health and Human Services and updated in 2018, cite adherence to the published guidelines as a protective factor against high LDL and total lipids in both adults and children.⁶ The guidelines for adults recommend 150 to 300 minutes of moderate-intensity or 75 to 150 minutes of vigorous-intensity aerobic exercise per week, as well as muscle-strengthening activities of moderate or greater intensity 2 or more days per week. Brisk walking is included as an example of a moderate-intensity activity. These same guidelines are cited and endorsed by the American College of Sports Medicine and the American Heart Association.^7,8

Editor’s takeaway

The lipid reductions achieved from walking—if any—are minimal. By themselves, these small reductions will not accomplish our ­lipid-lowering goals. However, cholesterol goals are primarily disease oriented. This evidence does not directly inform us of important patient-oriented outcomes, such as morbidity, mortality, and vitality.

Evidence summary

Walking’s impact on cholesterol levels is modest, inconsistent

A 2022 systematic review and meta-analysis of 21 studies (n = 1129) evaluated the effects of walking on lipids and lipoproteins in women older than 18 years who were overweight or obese and were not taking any lipid-­lowering medications. Median TC was 206 mg/dL and median LDL was 126 mg/dL.¹

The primary outcome found that walking decreased TC and LDL levels independent of diet and weight loss. Twenty studies reported on TC and showed that walking significantly decreased TC levels compared to the control groups (raw mean difference [RMD] = 6.7 mg/dL; 95% CI, 0.4-12.9; P = .04). Fifteen studies examined LDL and showed a significant decrease in LDL levels with walking compared to control groups (RMD = 7.4 mg/dL; 95% CI, 0.3-14.5; P = .04). However, the small magnitude of the changes may have little clinical impact.¹

There were no significant changes in the walking groups compared to the control groups for triglycerides (17 studies; RMD = 2.2 mg/dL; 95% CI, –8.4 to 12.8; P = .68) or high-density lipoprotein (HDL) (18 studies; RMD = 1.5 mg/dL; 95% CI, –0.4 to 3.3; P = .12). Included studies were required to be controlled but were otherwise not described. The overall risk for bias was determined to be low.¹

A 2020 RCT (n = 22) assessed the effects of a walking intervention on cholesterol and cardiovascular disease (CVD) risk in individuals ages 40 to 65 years with moderate CVD risk but without diabetes or CVD.² Moderate CVD risk was defined as a 2% to 5% ­10-year risk for a CVD event using the European HeartScore, which incorporates age, sex, blood pressure, lipid levels, and smoking status³; however, study participants were not required to have hyperlipidemia. Participants were enrolled in a 12-week, nurse-led intervention of moderate-paced walking for 30 to 45 minutes 5 times weekly.

Individuals in the intervention group had significant decreases in average TC levels from baseline to follow-up (244.6 mg/dL vs 213.7 mg/dL; P = .001). As a result, participants’ average 10-year CVD risk was significantly reduced from moderate risk to low risk (2.6% vs 1.8%; P = 038) and was significantly lower in the intervention group than in the control group at follow-up (1.8% vs 3.1%; P = .019). No blinding was used, and the use of lipid-lowering medications was not reported, which could have impacted the results.²

A 2008 RCT (n = 67) examined the effect of a home-based walking program (12 weeks of brisk walking, at least 30 min/d and at least 5 d/wk, with at least 300 kcal burned per walk) vs a sedentary control group in men ages 45 to 65 years with hyperlipidemia (TC > 240 mg/dL and/or TC/­HDL-C ratio ≥ 6) who were not receiving lipid-lowering medication. There were no significant changes from baseline to follow-up in the walking group compared to the control group in TC (adjusted mean difference [AMD] = –9.3 mg/dL; 95% CI, –22.8 to 4.64; P = .19), HDL-C (AMD = 2.7 mg/dL; 95% CI, –0.4 to 5.4; P = .07) or triglycerides (AMD = –26.6 mg/dL; 95% CI, –56.7 to 2.7; P = .07).⁴

The lipid reductions achieved from walking—if any—are minimal.

A 2002 RCT (n = 111) of sedentary men and women (BMI, 25-35; ages, 40-65 years) with dyslipidemia (LDL of 130-190 mg/dL, or HDL < 40 mg/dL for men or < 45 mg/dL for women) examined the impact of various physical activity levels for 8 months when compared to a control group observed for 6 months. The group assigned to low-amount, moderate-intensity physical activity walked an equivalent of 12 miles per week.⁵

Continue to: In this group...

In this group, there was a significant decrease in average triglyceride concentrations from baseline to follow-up (mean ± standard error = 196.8 ± 30.5 mg/dL vs 145.2 ± 16.0 mg/dL; P < .001). Significance of the change compared with changes in the control group was not reported, although triglycerides in the control group increased from baseline to follow-up (132.1 ± 11.0 vs 155.8 ± 14.9 mg/dL). There were no significant changes from baseline to follow-up in TC (194 ± 4.8 vs 197.9 ± 5.4 mg/dL), LDL (122.7 ± 4.0 vs 127.8 ± 4.1 mg/dL), or HDL (42.0 ± 1.9 vs 43.1 ± 2.5 mg/dL); P values of pre-post changes and comparison to control group were not reported.⁵

Recommendations from others

The Physical Activity Guidelines for Americans, published by the Department of Health and Human Services and updated in 2018, cite adherence to the published guidelines as a protective factor against high LDL and total lipids in both adults and children.⁶ The guidelines for adults recommend 150 to 300 minutes of moderate-intensity or 75 to 150 minutes of vigorous-intensity aerobic exercise per week, as well as muscle-strengthening activities of moderate or greater intensity 2 or more days per week. Brisk walking is included as an example of a moderate-intensity activity. These same guidelines are cited and endorsed by the American College of Sports Medicine and the American Heart Association.^7,8

Editor’s takeaway

The lipid reductions achieved from walking—if any—are minimal. By themselves, these small reductions will not accomplish our ­lipid-lowering goals. However, cholesterol goals are primarily disease oriented. This evidence does not directly inform us of important patient-oriented outcomes, such as morbidity, mortality, and vitality.

Evidence summary

Walking’s impact on cholesterol levels is modest, inconsistent

A 2022 systematic review and meta-analysis of 21 studies (n = 1129) evaluated the effects of walking on lipids and lipoproteins in women older than 18 years who were overweight or obese and were not taking any lipid-­lowering medications. Median TC was 206 mg/dL and median LDL was 126 mg/dL.¹

The primary outcome found that walking decreased TC and LDL levels independent of diet and weight loss. Twenty studies reported on TC and showed that walking significantly decreased TC levels compared to the control groups (raw mean difference [RMD] = 6.7 mg/dL; 95% CI, 0.4-12.9; P = .04). Fifteen studies examined LDL and showed a significant decrease in LDL levels with walking compared to control groups (RMD = 7.4 mg/dL; 95% CI, 0.3-14.5; P = .04). However, the small magnitude of the changes may have little clinical impact.¹

There were no significant changes in the walking groups compared to the control groups for triglycerides (17 studies; RMD = 2.2 mg/dL; 95% CI, –8.4 to 12.8; P = .68) or high-density lipoprotein (HDL) (18 studies; RMD = 1.5 mg/dL; 95% CI, –0.4 to 3.3; P = .12). Included studies were required to be controlled but were otherwise not described. The overall risk for bias was determined to be low.¹

A 2020 RCT (n = 22) assessed the effects of a walking intervention on cholesterol and cardiovascular disease (CVD) risk in individuals ages 40 to 65 years with moderate CVD risk but without diabetes or CVD.² Moderate CVD risk was defined as a 2% to 5% ­10-year risk for a CVD event using the European HeartScore, which incorporates age, sex, blood pressure, lipid levels, and smoking status³; however, study participants were not required to have hyperlipidemia. Participants were enrolled in a 12-week, nurse-led intervention of moderate-paced walking for 30 to 45 minutes 5 times weekly.

Individuals in the intervention group had significant decreases in average TC levels from baseline to follow-up (244.6 mg/dL vs 213.7 mg/dL; P = .001). As a result, participants’ average 10-year CVD risk was significantly reduced from moderate risk to low risk (2.6% vs 1.8%; P = 038) and was significantly lower in the intervention group than in the control group at follow-up (1.8% vs 3.1%; P = .019). No blinding was used, and the use of lipid-lowering medications was not reported, which could have impacted the results.²

A 2008 RCT (n = 67) examined the effect of a home-based walking program (12 weeks of brisk walking, at least 30 min/d and at least 5 d/wk, with at least 300 kcal burned per walk) vs a sedentary control group in men ages 45 to 65 years with hyperlipidemia (TC > 240 mg/dL and/or TC/­HDL-C ratio ≥ 6) who were not receiving lipid-lowering medication. There were no significant changes from baseline to follow-up in the walking group compared to the control group in TC (adjusted mean difference [AMD] = –9.3 mg/dL; 95% CI, –22.8 to 4.64; P = .19), HDL-C (AMD = 2.7 mg/dL; 95% CI, –0.4 to 5.4; P = .07) or triglycerides (AMD = –26.6 mg/dL; 95% CI, –56.7 to 2.7; P = .07).⁴

The lipid reductions achieved from walking—if any—are minimal.

A 2002 RCT (n = 111) of sedentary men and women (BMI, 25-35; ages, 40-65 years) with dyslipidemia (LDL of 130-190 mg/dL, or HDL < 40 mg/dL for men or < 45 mg/dL for women) examined the impact of various physical activity levels for 8 months when compared to a control group observed for 6 months. The group assigned to low-amount, moderate-intensity physical activity walked an equivalent of 12 miles per week.⁵

Continue to: In this group...

In this group, there was a significant decrease in average triglyceride concentrations from baseline to follow-up (mean ± standard error = 196.8 ± 30.5 mg/dL vs 145.2 ± 16.0 mg/dL; P < .001). Significance of the change compared with changes in the control group was not reported, although triglycerides in the control group increased from baseline to follow-up (132.1 ± 11.0 vs 155.8 ± 14.9 mg/dL). There were no significant changes from baseline to follow-up in TC (194 ± 4.8 vs 197.9 ± 5.4 mg/dL), LDL (122.7 ± 4.0 vs 127.8 ± 4.1 mg/dL), or HDL (42.0 ± 1.9 vs 43.1 ± 2.5 mg/dL); P values of pre-post changes and comparison to control group were not reported.⁵

Recommendations from others

The Physical Activity Guidelines for Americans, published by the Department of Health and Human Services and updated in 2018, cite adherence to the published guidelines as a protective factor against high LDL and total lipids in both adults and children.⁶ The guidelines for adults recommend 150 to 300 minutes of moderate-intensity or 75 to 150 minutes of vigorous-intensity aerobic exercise per week, as well as muscle-strengthening activities of moderate or greater intensity 2 or more days per week. Brisk walking is included as an example of a moderate-intensity activity. These same guidelines are cited and endorsed by the American College of Sports Medicine and the American Heart Association.^7,8

Editor’s takeaway

The lipid reductions achieved from walking—if any—are minimal. By themselves, these small reductions will not accomplish our ­lipid-lowering goals. However, cholesterol goals are primarily disease oriented. This evidence does not directly inform us of important patient-oriented outcomes, such as morbidity, mortality, and vitality.

Evidence summary

Mixed results on exercise’s effect on neuropsychiatric symptoms

A 2020 systematic review and meta-analysis of 18 RCTs investigated the effect of home-based physical activity on several markers of behavioral and psychological symptoms of dementia (BPSD). These symptoms were measured using the caregiver-completed neuropsychiatric inventory (NPI), which in­cludes agitation. There was substantial heterogeneity between trials; however, 4 RCTs (472 patients) were included in a meta-­analysis of the NPI. These RCTs were nonblinded, given the nature of the intervention.¹

Interventions to enhance physical activity ranged from 12 weeks to 2 years in duration, with 2 to 8 contacts from the study team per week. The type of physical activity varied and included cardiorespiratory endurance, balance training, resistance training, and activities of daily living training.¹

Exercise was associated with significantly fewer symptoms on the NPI, although the effect size was small (standard mean difference [SMD] = –0.37; 95% CI, –0.57 to –0.17). Heterogeneity in the interventions and assessments were limitations to this meta-analysis.¹

A 2015 systematic review and meta-­analysis of 18 RCTs compared the effect of exercise interventions against a control group for the treatment of BPSD, utilizing 10 behavioral and 2 neurovegetative components of the NPI (each scored from 0 to 5) in patients with dementia. Studies were included if they used ≥ 1 exercise intervention compared to a control or usual care group without additional exercise recommendations. Thirteen studies had a multicomponent training intervention (≥ 2 exercise types grouped together in the same training session), 2 used tai chi, 4 used walking, and 1 used dance and movement therapy. These RCTs were conducted in a variety of settings, including community-dwelling and long-term care facilities (n = 6427 patients).²

Exercise did not reduce global BPSD (N = 4441 patients), with a weighted mean difference (WMD) of −3.9 (95% CI, −9.0 to 1.2; P = .13). Exploratory analysis did not show improvement in aberrant motor behavior with exercise (WMD = –0.55; 95% CI, –1.10 to 0.001; P = .05). Limitations of this review included the small number of studies, heterogeneity of the population, and limitations in data accessibility.²

A 2017 hospital-based RCT evaluated the effects of a short-term exercise program on neuropsychiatric signs and symptoms in patients with dementia in 3 specialized dementia care wards (N = 85). Patients had a diagnosis of dementia, minimum length of stay of 1 week, no delirium, and the ability to perform the Timed Up and Go Test. The intervention group included a 2-week exercise program of four 20-minute exercise sessions per day on 3 days per week, involving strengthening or endurance exercises, in addition to treatment as usual. The control group included a 2-week period of social-stimulation programs consisting of table games for 120 minutes per week, in addition to treatment as usual.³

Exercise remains a small tool to address a big problem.

Of 85 patients randomized, 15 (18%) were lost to follow-up (14 of whom were discharged early from the hospital). Among the 70 patients included in the final analysis, the mean age was 80 years; 47% were female and 53% male; and the mean Mini-Mental Status Examination score was 18.3 (≤ 23 indicates dementia). In both groups, most patients had moderate dementia, moderate neuropsychiatric signs and symptoms, and a low level of psychotic symptoms. Patients in the intervention group had a higher adherence rate compared with those in the control group.³

Continue to: The primary outcome...

The primary outcome was neuropsychiatric signs and symptoms as measured by the Alzheimer’s Disease Cooperative Study–­Clinical Global Impression of Change (ADCS-CGIC). Compared to the control group, the intervention group experienced greater improvement on the ADCS-CGIC dimensions of emotional agitation (SMD = –0.9; P < .001), lability (SMD = –1.1; P < .001), psychomotor agitation (SMD = –0.7; P = .01), and verbal aggression (SMD = –0.5; P = .04). However, there were no differences between groups in the physical aggression dimension. Trial limitations included potential impact of the drop-out rate and possible blinding issues, as nursing staff performing assessments could have seen to which group a patient was allocated.³

A 2016 factorial cluster RCT of 16 nursing homes (with at least 60% of the population having dementia) compared the use of ­person-centered care vs person-centered care plus at least 1 randomly assigned additional intervention (eg, antipsychotic medication use review, social interaction interventions, and exercise over a period of 9 months) (n = 277, with 193 analyzed per protocol). Exercise was implemented at 1 hour per week or at an increase of 20% above baseline and compared with a control group with no change in exercise.⁴

Exercise significantly improved neuropsychiatric symptoms. The baseline NPI score of 14.54 improved by –3.59 (95% CI, –7.08 to –0.09; P < .05). However, none of the study interventions significantly improved the agitation-specific scores. The primary limitation of this study was that antipsychotic prescribing was at the discretion of the provider and not according to a protocol. In addition, the authors noted that the trial was inadequately powered to correct for testing 3 primary outcomes.⁴

Editor’s takeaway

Dementia and dementia with agitation are challenging conditions to treat. Disappointingly, physical exercise had inconsistent and generally minimal effect on agitation in dementia. Nevertheless, exercise had other positive effects. So, considering the benefits that exercise does provide, its low cost, and its limited adverse effects, exercise remains a small tool to address a big problem.

Evidence summary

Mixed results on exercise’s effect on neuropsychiatric symptoms

A 2020 systematic review and meta-analysis of 18 RCTs investigated the effect of home-based physical activity on several markers of behavioral and psychological symptoms of dementia (BPSD). These symptoms were measured using the caregiver-completed neuropsychiatric inventory (NPI), which in­cludes agitation. There was substantial heterogeneity between trials; however, 4 RCTs (472 patients) were included in a meta-­analysis of the NPI. These RCTs were nonblinded, given the nature of the intervention.¹

Interventions to enhance physical activity ranged from 12 weeks to 2 years in duration, with 2 to 8 contacts from the study team per week. The type of physical activity varied and included cardiorespiratory endurance, balance training, resistance training, and activities of daily living training.¹

Exercise was associated with significantly fewer symptoms on the NPI, although the effect size was small (standard mean difference [SMD] = –0.37; 95% CI, –0.57 to –0.17). Heterogeneity in the interventions and assessments were limitations to this meta-analysis.¹

A 2015 systematic review and meta-­analysis of 18 RCTs compared the effect of exercise interventions against a control group for the treatment of BPSD, utilizing 10 behavioral and 2 neurovegetative components of the NPI (each scored from 0 to 5) in patients with dementia. Studies were included if they used ≥ 1 exercise intervention compared to a control or usual care group without additional exercise recommendations. Thirteen studies had a multicomponent training intervention (≥ 2 exercise types grouped together in the same training session), 2 used tai chi, 4 used walking, and 1 used dance and movement therapy. These RCTs were conducted in a variety of settings, including community-dwelling and long-term care facilities (n = 6427 patients).²

Exercise did not reduce global BPSD (N = 4441 patients), with a weighted mean difference (WMD) of −3.9 (95% CI, −9.0 to 1.2; P = .13). Exploratory analysis did not show improvement in aberrant motor behavior with exercise (WMD = –0.55; 95% CI, –1.10 to 0.001; P = .05). Limitations of this review included the small number of studies, heterogeneity of the population, and limitations in data accessibility.²

A 2017 hospital-based RCT evaluated the effects of a short-term exercise program on neuropsychiatric signs and symptoms in patients with dementia in 3 specialized dementia care wards (N = 85). Patients had a diagnosis of dementia, minimum length of stay of 1 week, no delirium, and the ability to perform the Timed Up and Go Test. The intervention group included a 2-week exercise program of four 20-minute exercise sessions per day on 3 days per week, involving strengthening or endurance exercises, in addition to treatment as usual. The control group included a 2-week period of social-stimulation programs consisting of table games for 120 minutes per week, in addition to treatment as usual.³

Exercise remains a small tool to address a big problem.

Of 85 patients randomized, 15 (18%) were lost to follow-up (14 of whom were discharged early from the hospital). Among the 70 patients included in the final analysis, the mean age was 80 years; 47% were female and 53% male; and the mean Mini-Mental Status Examination score was 18.3 (≤ 23 indicates dementia). In both groups, most patients had moderate dementia, moderate neuropsychiatric signs and symptoms, and a low level of psychotic symptoms. Patients in the intervention group had a higher adherence rate compared with those in the control group.³

Continue to: The primary outcome...

The primary outcome was neuropsychiatric signs and symptoms as measured by the Alzheimer’s Disease Cooperative Study–­Clinical Global Impression of Change (ADCS-CGIC). Compared to the control group, the intervention group experienced greater improvement on the ADCS-CGIC dimensions of emotional agitation (SMD = –0.9; P < .001), lability (SMD = –1.1; P < .001), psychomotor agitation (SMD = –0.7; P = .01), and verbal aggression (SMD = –0.5; P = .04). However, there were no differences between groups in the physical aggression dimension. Trial limitations included potential impact of the drop-out rate and possible blinding issues, as nursing staff performing assessments could have seen to which group a patient was allocated.³

A 2016 factorial cluster RCT of 16 nursing homes (with at least 60% of the population having dementia) compared the use of ­person-centered care vs person-centered care plus at least 1 randomly assigned additional intervention (eg, antipsychotic medication use review, social interaction interventions, and exercise over a period of 9 months) (n = 277, with 193 analyzed per protocol). Exercise was implemented at 1 hour per week or at an increase of 20% above baseline and compared with a control group with no change in exercise.⁴

Exercise significantly improved neuropsychiatric symptoms. The baseline NPI score of 14.54 improved by –3.59 (95% CI, –7.08 to –0.09; P < .05). However, none of the study interventions significantly improved the agitation-specific scores. The primary limitation of this study was that antipsychotic prescribing was at the discretion of the provider and not according to a protocol. In addition, the authors noted that the trial was inadequately powered to correct for testing 3 primary outcomes.⁴

Editor’s takeaway

Dementia and dementia with agitation are challenging conditions to treat. Disappointingly, physical exercise had inconsistent and generally minimal effect on agitation in dementia. Nevertheless, exercise had other positive effects. So, considering the benefits that exercise does provide, its low cost, and its limited adverse effects, exercise remains a small tool to address a big problem.

Evidence summary

Mixed results on exercise’s effect on neuropsychiatric symptoms

A 2020 systematic review and meta-analysis of 18 RCTs investigated the effect of home-based physical activity on several markers of behavioral and psychological symptoms of dementia (BPSD). These symptoms were measured using the caregiver-completed neuropsychiatric inventory (NPI), which in­cludes agitation. There was substantial heterogeneity between trials; however, 4 RCTs (472 patients) were included in a meta-­analysis of the NPI. These RCTs were nonblinded, given the nature of the intervention.¹

Interventions to enhance physical activity ranged from 12 weeks to 2 years in duration, with 2 to 8 contacts from the study team per week. The type of physical activity varied and included cardiorespiratory endurance, balance training, resistance training, and activities of daily living training.¹

Exercise was associated with significantly fewer symptoms on the NPI, although the effect size was small (standard mean difference [SMD] = –0.37; 95% CI, –0.57 to –0.17). Heterogeneity in the interventions and assessments were limitations to this meta-analysis.¹

A 2015 systematic review and meta-­analysis of 18 RCTs compared the effect of exercise interventions against a control group for the treatment of BPSD, utilizing 10 behavioral and 2 neurovegetative components of the NPI (each scored from 0 to 5) in patients with dementia. Studies were included if they used ≥ 1 exercise intervention compared to a control or usual care group without additional exercise recommendations. Thirteen studies had a multicomponent training intervention (≥ 2 exercise types grouped together in the same training session), 2 used tai chi, 4 used walking, and 1 used dance and movement therapy. These RCTs were conducted in a variety of settings, including community-dwelling and long-term care facilities (n = 6427 patients).²

Exercise did not reduce global BPSD (N = 4441 patients), with a weighted mean difference (WMD) of −3.9 (95% CI, −9.0 to 1.2; P = .13). Exploratory analysis did not show improvement in aberrant motor behavior with exercise (WMD = –0.55; 95% CI, –1.10 to 0.001; P = .05). Limitations of this review included the small number of studies, heterogeneity of the population, and limitations in data accessibility.²

A 2017 hospital-based RCT evaluated the effects of a short-term exercise program on neuropsychiatric signs and symptoms in patients with dementia in 3 specialized dementia care wards (N = 85). Patients had a diagnosis of dementia, minimum length of stay of 1 week, no delirium, and the ability to perform the Timed Up and Go Test. The intervention group included a 2-week exercise program of four 20-minute exercise sessions per day on 3 days per week, involving strengthening or endurance exercises, in addition to treatment as usual. The control group included a 2-week period of social-stimulation programs consisting of table games for 120 minutes per week, in addition to treatment as usual.³

Exercise remains a small tool to address a big problem.

Of 85 patients randomized, 15 (18%) were lost to follow-up (14 of whom were discharged early from the hospital). Among the 70 patients included in the final analysis, the mean age was 80 years; 47% were female and 53% male; and the mean Mini-Mental Status Examination score was 18.3 (≤ 23 indicates dementia). In both groups, most patients had moderate dementia, moderate neuropsychiatric signs and symptoms, and a low level of psychotic symptoms. Patients in the intervention group had a higher adherence rate compared with those in the control group.³

Continue to: The primary outcome...

The primary outcome was neuropsychiatric signs and symptoms as measured by the Alzheimer’s Disease Cooperative Study–­Clinical Global Impression of Change (ADCS-CGIC). Compared to the control group, the intervention group experienced greater improvement on the ADCS-CGIC dimensions of emotional agitation (SMD = –0.9; P < .001), lability (SMD = –1.1; P < .001), psychomotor agitation (SMD = –0.7; P = .01), and verbal aggression (SMD = –0.5; P = .04). However, there were no differences between groups in the physical aggression dimension. Trial limitations included potential impact of the drop-out rate and possible blinding issues, as nursing staff performing assessments could have seen to which group a patient was allocated.³

A 2016 factorial cluster RCT of 16 nursing homes (with at least 60% of the population having dementia) compared the use of ­person-centered care vs person-centered care plus at least 1 randomly assigned additional intervention (eg, antipsychotic medication use review, social interaction interventions, and exercise over a period of 9 months) (n = 277, with 193 analyzed per protocol). Exercise was implemented at 1 hour per week or at an increase of 20% above baseline and compared with a control group with no change in exercise.⁴

Exercise significantly improved neuropsychiatric symptoms. The baseline NPI score of 14.54 improved by –3.59 (95% CI, –7.08 to –0.09; P < .05). However, none of the study interventions significantly improved the agitation-specific scores. The primary limitation of this study was that antipsychotic prescribing was at the discretion of the provider and not according to a protocol. In addition, the authors noted that the trial was inadequately powered to correct for testing 3 primary outcomes.⁴

Editor’s takeaway

Dementia and dementia with agitation are challenging conditions to treat. Disappointingly, physical exercise had inconsistent and generally minimal effect on agitation in dementia. Nevertheless, exercise had other positive effects. So, considering the benefits that exercise does provide, its low cost, and its limited adverse effects, exercise remains a small tool to address a big problem.